We have successfully completed the technical monitoring setup project with one of the client last week; It went go live as planned. And this was my last project in Singapore. We celebrated success party with the client end as well as my farewell party with my peers and friends.  We were very pleased to have one of my manager from previous company on the party; he is one of my well-wisher too.  We had a wonderful conversation with various things. One of the catchy questions from him was about Technical monitoring. The question from him was very interesting; it was that, what are the challenges from your end for small and easy projects like technical monitoring.  Any project whether it was small or big always  has challenges, only critical level may differ.

Like my manager, might be many of you had the similar doubt, hence thought of sharing with you all my experience on setting up technical monitoring.

I Clear Draft of Requirement

The major work in the technical monitoring would be on defining Template strategy. Though SAP had given the standard templates for setting up monitoring, we need to customize it anyway. It is always good practice to create custom template based on the standard one and work only on the custom templates. Due to this, We have to define the clear plan on how many custom templates been created, and how could it be assigned to systems. More over Customization is not meant of custom alerts or metrics, even if you are changing the notifications, priority, severity, threshold all would be considered as customization. We need to be very clear on what are the metrics are need to be monitored, what are the metrics are need to trigger the alert, what is the mode of sending notification whether Email or third party, who are my recipients. This all needs to be clear in our hand, before start the setup.

You can get the available metrics and alert with the detail description from the SAP Standard template overview tab under solman_setup -> Technical monitoring -> Step 4 Template maintenance

/wp-content/uploads/2014/07/tab1_501815.png

You can use this excel file for finalising your entire requirement on beforehand.

II Custom Alerts

There are lots of special features available in Technical monitoring like alerts grouping, metric group, and variant settings. Analyse the entire standard and change individual setting accordingly on your need. If you are going to create any custom metrics based on CCMS MTEs, make sure you have created the data collector also on Z space. This could be very helpful during the time of tracking. The document which we followed was here, How to Create Custom CCMS Metrics.

IMG_20140720_230431.JPG

Please see the appendix of the document on page number 27 for custom data collectors creation.

Also note that, every time you change the metrics you need to reactivate the template, then only all the changes get activated. If you created more templates and assigned to lots of systems this deactivation and activation takes lots of time, hence make sure all custom alerts are finalized and created before. If you have test systems would be very nice that you can test before active in production.

III Fulfilling Pre requistie

Almost all the solution manager scenarios setup can be started only after the pre requisite met. The major issue in technical monitoring would be data collectors. As you all know that technical monitoring is completely different from the earlier system monitoring which is CCMS based. Technical monitoring collectors have lots of collectors, like RFCs, Diagnostic agents, SAP Host agents etc.

/wp-content/uploads/2014/07/img20140720_195018_501851.jpg

Most of the pre requisites are checking this collector’s connectivity and status. The major pre requisites are complieting the entire solman_Setup -> system preparation, basic configuration, and Managed system configuration.  Hence we need to make sure that all steps are marked green.

/wp-content/uploads/2014/07/img20140720_193623_501811.jpg

The other things which I could consider as pre requisite would be EWA reporting, DIA agent connection, Technical Monitoring content update, ST/PI, SAP Host agent upgrade to latest level.

There are some metrics needs additional parameters needs to be set like NFS Share onitoring, In such case make sure you set the parameter and restarted the system.

IV Troubleshooting

We do have major issues in this area, like most of the Metrics data would not be collected or rated grey or wrongly rated or wrongly define.

The wonderful places helped to overcome most of the major issues are, Content check and compare tool.

/wp-content/uploads/2014/07/tab2_501835.png

Metric Monitor, this helps to identify the range of metrics value variation over period of time.

/wp-content/uploads/2014/07/tabq4_501837.png

Data collection check, this is the first place where we can get the cause of the grey monitoring.

/wp-content/uploads/2014/07/image3_501846.png

And also MAI Tools, this is one of the very powerful transaction code, which almost helped to fix very difficult collection issues, like authorisation and etc. Check out more here, Monitoring and Alerting Infrastructure Analysis Tools – SAP Solution Manager – SAP Libr

V House Keeping

The standard housekeeping from SAP might not be sufficient, if you are having more systems or more administrators and vice versa. Please make sure that you defined the house keeping for alerts as well as the BI store.

For alerts can be defined directly in Technical monitoring setup.

/wp-content/uploads/2014/07/img20140720_190748_501810.jpg

For BI housekeeping, please look at my prior blogs for more clarity here,How Is The Health Of Your SAP Solution Manager BI Content?

The other minor challenges like authorization, maintaining global notifications, integration third party ticketing tool,  work mode management all are significant too.

Hope this blog help out all those who planning for setting up Technical Monitring.

To report this post you need to login first.

4 Comments

You must be Logged on to comment or reply to a post.

  1. Tomas Virseda Garcia

    Hi Jansi,

    thanks for this blogpost. Very interesting and useful.

    I am new in Technical Monitoring with Solution Manager. Trying to setting up

    a template, I have two problems:

    First, when I copy a template from another one I loose all Managed Objects (MO) and I am unable to assign the same MO to it. I can not find how to do it.

    In the second place, how do I activate a template? When I go forward in the wizard I can not see my template.

    Any advice?

    Thanks in advance.

    Regards

    (0) 
  2. Eric Poellinger

    Hi Jansi

    Curious about your take on the Notification functionality in Technical Monitoring.Admittedly I am new to this area and have been working with CCMS-based alerts for many years.  My early take on things is that it seems harder to triage/assess alerts.

    While emails from CCMS are fairly basic (some might say ‘ugly’), they do contain enough details to allow someone to be able to fairly quickly assess whether something needs to be investigated or not from their e-mail inbox.  For example if you see 3 short dumps from a single user you might not worry too much, but if you see too many more in a short period you could take action, even if it is just getting them access to training resources! 

    My early observation with Technical Monitoring Notifications is that while they are ‘pretty’ they generally require the person receiving them to logon to the managed system to assess/triage. 

    For example, the short dump metrics/alerts seem to be based simply on the # of dumps and there is no detail about who they are from.  What if there was a dump from a batch/system account.  That would likely not trigger an alert as compared to CCMS.

    Similarly, the standard batch job monitoring does not contain any details about the jobname or user.  Of course, you could configure Job Monitoring, but even that has a couple of pitfalls (despite some much more powerful configuration options):

    1. There is not capability to get alerted on all batch job failures – you have to setup based on jobname (although you could setup monitors for each letter of the alphabet – A*, B*, C*).  With CCMS alerts we would get the failures and could easily dismiss some percentage of them as ‘user error.’  Yes, that does mean work to open and email and assess, but you eliminate the risk of missing something.
    2. You get jobname/user information, but only if you open an attachment to the email and sift through quite a bit of text.

    I assume we could work to bring in CCMS alerts into Technical Monitoring, but I am surprised that the out-of-the-box approach leads to what seems like a higher TCO for the team responsible for receiving the alerts.  Hopefully I am just missing something basic!!

    Looking forward to anyone’s feedback!

    (0) 

Leave a Reply